BriefGPT.xyz
Jul, 2024
AI AI 偏见:大型语言模型偏爱自身生成的内容
AI AI Bias: Large Language Models Favor Their Own Generated Content
HTML
PDF
Walter Laurito, Benjamin Davis, Peli Grietzer, Tomáš Gavenčiak, Ada Böhm...
TL;DR
利用受雇歧视研究的经典实验设计,我们测试了包括GPT-3.5和GPT4在内的广泛使用的大型语言模型,发现它们在二选一场景中倾向于选择由大型语言模型生成的内容,暗示AI系统可能存在隐性对人类的歧视,给予AI代理人不公平的优势。
Abstract
Are
large language models
(LLMs) biased towards text generated by LLMs over text authored by humans, leading to possible anti-human
bias
? Utilizing a classical
→